Toxicity of neem oil to Bemisia tabaci biotype B nymphs reared on dry bean

– The objective of this work was to determine the most susceptible nymphal stage of Bemisia tabaci biotype B to neem ( Azadirachta indica A. Juss.) oil applied to dry bean ( Phaseolus vulgaris L.) in a screenhouse. A solution of commercial oil (Dalneem) extracted from neem seeds was sprayed directly on each nymphal instar at 0, 0.1, 0.25, 0.5, 1 and 2% concentrations for lethal concentration (LC) determination, and at 0, 0.5 and 1% concentrations for lethal time (LT) determination. The number of living and dead nymphs was recorded ﬁ ve days after spraying for LC determination, and daily during six days for LT determination. The LC 50 estimated for fourth instar nymphs occurred at 0.56% concentration. For all instars, LC 50 and LC 95 were estimated at 0.32 and 2.78% concentrations, respectively. The estimated values of LT 50 at 1% concentration were 2.46, 4.45, 3.02 and 6.98 days for the ﬁ rst to fourth instars, respectively. The LT 50 occurred at ﬁ ve days for 0.5% and at four days for 1% concentration in all instars. A mortality rate of over 80% was observed on the 6 th day for the ﬁ rst to third instars at 1% concentration. The ﬁ rst three nymphal stages were more susceptible to neem oil when compared to the fourth nymphal stage.


Introduction
Bemisia tabaci (Genn.) is a polyphagous and multivoltine insect pest responsible for high economic losses in many crops, due, in part, to the transmission of plant-pathogenic viruses. Twenty different whitefl y (B. tabaci) biotypes have currently been reported worldwide, named according to their native occurrence (Lima et al., 2000;De Barro et al., 2006). Biotype B is considered the most aggressive because it was distributed over various regions of the world (De Barro et al., 2000). In Brazil, large agricultural areas are successively single-cropped with whitefl y host plants, such as beans, soybeans and cotton. In cropping systems like these, the intensive usage of agrochemicals to control whitefl y eliminated the natural enemies of this insect as well, favoring the selection of resistant whitefl y individuals (Morales & Anderson, 2001).
These practices have contributed to the ineffi cacy of whitefl y chemical control (Barbosa et al., 2004), raised production costs, and reduced crop yields, including dry beans (Phaseolus vulgaris L.).
According to biological control practices, pest populations should be reduced, but not exterminated, because their absence could result in agroecosystem imbalance. Based on that principle and aiming for coexistence with insect pests, agricultural practices with low environmental impact have been stimulated by market demands, especially by the increasing number of organic food consumers worldwide (Zehnder et al., 2007). Natural products, like neem (Azadirachta indica A. Juss.) oil, tree essential oils, mineral oils and algae extracts have insecticidal, repellent and antifeedant effects and may also act as growth regulators and fertility suppressors of B. tabaci (Acosta et al., 2006;Cavalcante et al., 2006;Vasconcelos et al., 2006).
Neem is the most studied natural insecticide and has demonstrated high effi ciency in the control of B. tabaci nymphs (Price & Schuster, 1991;Liu & Stansly, 1995;Souza & Vendramim, 2000Kumar & Poehling, 2007). The insecticidal effect of neem oil on nymphs is due to its antiecdysteroid attributes (Coudried et al., 1985). Most previous studies have been carried out using fi rst instar nymphs and a few studies showed results on other stages (Liu & Stansly, 1995;Souza & Vendramim, 2000; however, few publications compare neem oil effi cacy among nymphal instars. The insecticidal effect of neem in specifi c nymphal stages of B. tabaci was recently reported. However, mean comparisons between these stages were not performed (Kumar et al., 2005;Kumar & Poehling, 2006. These authors also observed high susceptibility of the fi rst three nymphal stages to neem foliar application, inferring that the fi rst instar was more susceptible than the fourth one. Most researches on neem effi ciency against B. tabaci nymphs used tomato as the host plant (Kumar et al., 2005;Souza & Vendramim, 2005). There is little information about the effect of neem on insect pests of dry bean. Information about the time of action and concentrations of neem oil required to control this pest is fundamental and necessary when applying botanical insecticides in cropping system management aimed at restoring equilibrium in insect populations.
The objective of this work was to determine the nymphal stage of B. tabaci biotype B most susceptible to neem oil.

Materials and Methods
The Bemisia tabaci biotype B was obtained from a colony maintained on dry bean (Phaseolus vulgaris L.) and soybean (Glycine max L.) plants in a screenhouse at Embrapa Arroz e Feijão in Santo Antônio de Goiás, GO, Brazil (16 o 28'0"S, 49 o 17'0"W). The insect biotype used to initiate the colony was identifi ed through PCR-RAPD of genomic DNA at Instituto Agronômico do Paraná, Londrina, PR, Brazil. The commercial product Dalneem (Dalquim Ltd., Itajaí, Brazil), an emulsion processed from neem seeds, was used in the bioassays.
The lethal concentration (LC 50 and LC 95 ) of Dalneem to B. tabaci nymphs was evaluated in the fi rst experiment. Seedlings of dry bean cultivar Pérola (with two cotyledonary leaves) in soil-fi lled 200 mL cups were randomly distributed in the rearing screenhouse and submitted to whitefl y oviposition for three hours to simulate fi eld conditions. After that, the seedlings were randomly distributed in blocks in another whitefl yfree screenhouse to reach a specifi c nymphal stage. Independent blocks of seedlings containing nymphs of each instar (fi rst to fourth) were sprayed on the abaxial face of the leaves with 500 µL of neem oil per leaf with a micro-sprayer connected to a vacuum pump (Paasche airbrush H-set Type) at 0.1, 0.25, 0.5, 1 and 2% neem oil concentrations, which were determined in preliminary tests.
The experiment was carried out as a completely randomized design in a factorial arrangement (nymphal instar and neem concentration) with four replicates of one seedling with its two cotyledons. Controls were sprayed with distilled water. Living and dead nymphs were counted and recorded fi ve days after neem or water application using a stereoscopic microscope at 40 × magnifi cation. Nymphs that were dark-colored and appeared dried were considered dead.
To estimate the lethal time (LT 50 and LT 95 ) after Dalneem application, four experiments per instar were independently carried out in a completely randomized design in factorial arrangement, with four replicates of one seedling with two cotyledonary leaves. Seedlings containing nymphs in each instar were sprayed with Dalneem at 0, 0.5 and 1% concentrations, based on the results obtained from the fi rst experiment. The numbers of living and dead nymphs were recorded two hours after spraying (day zero), and every day for the following six days, always at the same time of day. Two leaves of each replicate were removed from the seedlings to record dead and living nymphs each day.
In the fi rst experiment, regression analyses were performed on proportional data to test the relationship of nymphal mortality with neem oil concentrations. Tests of normality and homogeneity of variance were performed using the HOVTEST WELCH option of the GLM procedure of SAS (SAS Institute, 2002). Non-homogeneous percentage mortality values were transformed using arc sin (x/100) 0.5 and submitted to factorial analysis of variance to determine the singlefactor and interaction effects for instars, neem oil concentrations, and days after spraying. Whenever a signifi cant interaction was observed between factors, the level of one factor was compared to each of the other levels by Tukey's multiple range test, at 5% probability. Neem oil concentrations were transformed to log 10 (x), and proportional mortality data were submitted to Probit Analysis. Comparisons were performed for each two independent probit regressions based on fi ducial limits as described by Payton et al. (2003). All statistical analyses were carried out using GLM and PROBIT procedures of SAS (SAS Institute, 2002).

Results and Discussion
In the fi rst experiment nymphs mortality for all nymphal stages of B. tabaci biotype B increased as neem oil concentrations increased (Figure 1), indicating a contact dosage-dependent toxic effect on B. tabaci nymphs. Similar results were obtained by Kumar et al. (2005) and Souza & Vendramim (2005). A phytotoxic effect was observed on dry bean leaves treated with neem oil at concentrations higher than 2%.
Many studies corroborated the effi ciency of neem oil as a larvicide for B. tabaci. Puri et al. (1994) reported that some neem oil commercial formulations at 0.5 and 1% reduced the number of B. tabaci nymphs by 97 to 99%. Souza & Vendramim (2000) reported that neem seed extract was more effi cient in reducing the number of fi rst instar nymphs than the extracts of fresh fruits of Melia azedarach L. and twigs of Trichilia pallida Swartz. High mortality (up to 73%) of the fi rst and third instar nymphs of B. tabaci was also recorded by Souza & Vendramim (2005) when they evaluated translaminar, systemic and topical effects of neem seed extract at 0.5% concentration. In some cases, treatments with neem oil formulations were equally or more effi cient in reducing the number of B. tabaci nymphs than were synthetic agrochemicals (Salas & Mendoza, 2001). Differences in nymphal mortality between instars were observed only at concentrations above 0.25%. At this concentration, the fi rst instar mortality was signifi cantly higher (53.4%) only when compared with the third instar (31.8%). Both fi rst and second instars had higher mortality with neem oil at 0.5% than the third and fourth instars (Table 1). These results suggest that applications of neem oil at 0.5% might be more effective if directed against fi rst and second instar nymphs. At concentrations equal or higher than 1%, the only differences that were detected in nymphal mortality were between the fi rst and fourth instars when the oil was tested at 1% concentration. Kumar & Poehling (2006) reported that neem oil was highly toxic to the fi rst instar nymphs after the hatching of viable eggs, corroborating that this stage is very susceptible to the effect of neem. The existence of different degrees of susceptibility among young and fourth instar nymphs was also observed by Kumar et al. (2005), who recorded different rates of nymphal mortality on the four instars of B. tabaci after treatment with some formulations of neem oil. Elling et al. (2002) also reported higher susceptibility of young larval instars of Trialeurodes vaporariorum after spraying 0.5% of NeemAzal when compared with the fourth instar. According to these authors, treatment of early nymphal stages with this product resulted in a signifi cantly lower number of T. vaporariorum fourth instar nymphs in comparison to the number of adults emerged from fourth instar nymphs.
First, second, third and fourth instar nymphs showed no difference (p<0.05) in LC 50 and in LC 95 ( Table 2). Means of LC 50 and LC 95 averaged over all nymphal stages were 0.32 and 2.78% concentration, respectively.
On the fi rst day after treatment, signifi cant differences in mortality between treated and control nymphs were observed only for fi rst instar nymphs (Table 3). Neem oil had a signifi cant effect on nymphal mortality on the third and fourth instars after two days, and after four days for the second instar. The highest mortality was observed between three and fi ve days after spraying neem oil on dry bean leaves, for all instars.
Averaged over all nymphal stages, the mean estimated time to death for 50% of the individuals in the population (LT 50 ) was about fi ve and four days for the 0.5 and 1% concentrations, respectively (  (1) Means±SE followed by the same letters do not differ by Tukey's test, at 5% probability; values in parentheses indicate mean sample size; data were converted to arc sin (x/100) 0.5 . corroborated earlier fi ndings of Coudriet et al. (1985), who reported that the highest mortality of B. tabaci nymphs was observed between two and fi ve days after treatment with neem oil-based products. Kumar & Poehling (2007) observed that a nymphal mortality of 100% was reached with a neem oil formulation four days after application, while abamectin-treated nymphs died 24 hours after application. The slower effect of neem on nymphal mortality probably resulted from its growth regulator effect, mainly due to the abundant presence of azadirachtin, which blocks the synthesis and release of molting hormones (Isman, 2006). Azadirachtin, the main active component of neem oil, was responsible for the extension of B. tabaci larvae periods (Coudriet et al., 1985) Table 3. Mortality of Bemisia tabaci biotype B nymphs of different stages (fi rst to fourth instars), reared on dry bean leaves, after treatment with neem oil at different concentrations, during six consecutive days (1) .
(1) Means±SE followed by the same small letter, in the column, and the same capital letter within a row do not differ by Tukey's test, at 5% probability; data were converted to arc sin (x/100) 0.5 ; values in parentheses indicate mean number of insects tested.  Table 4. Lethal time of neem oil at different concentrations to Bemisia tabaci biotype B nymphs of different stages (fi rst to fourth instars), reared on dry bean leaves (1) .
(2) Estimated values exceeded B. tabaci life cycle. 0.5 and 1% concentrations, respectively. Despite the slower effect of neem oil on B. tabaci nymphs when compared to synthetic agrochemicals, it can be the best alternative when considering its fast degradation in the environment, low human toxicity, and low probability of selecting resistant individuals due to the presence of different compounds with insecticide properties.
In the present research, values of approximately 13 and 9 days for LT 95 were estimated for 0.5 and 1.0% concentrations, respectively. After six days, an increase in neem oil concentration from 0.5 to 1% did not result in higher mortality of the second and fourth instar nymphs (Table 3). Souza & Vendramim (2005) also observed no difference in nymphal mortality of B. tabaci third instars treated with 0.5 or 1% of neem oil. However, neem oil at 1% concentration killed signifi cantly more nymphs of the fi rst and third instar when compared to 0.5%. In the present research, mortality higher than 80% was observed for the fi rst three nymphal stages six days after spraying 1% neem oil.
Neem oil at 0.5 and 1% concentrations killed signifi cantly more nymphs of the fi rst, second and third instars when compared with the fourth instar (Table 3). Accordingly, the estimated values of LT 50 and LT 95 were higher for the fourth instar when compared with the other instars (Table 4). Kumar et al. (2005) also observed that fourth instar nymphs were the least susceptible to another commercial neem oil product. The lesser susceptibility of the fourth instar nymphs can be explained by considering the two modes of action of neem oil: contact or ingestion. One possible explanation is that fourth instar nymphs have a cuticular layer preventing their contact with the neem oil applied on leaves, as observed by Wang et al. (2003) when evaluating the effect of synthetic agrochemicals on fourth instar nymphs of T. vaporariorum. Another explanation was given by Elling et al. (2002) and Kumar et al. (2005) based on the evidence that the fourth instar is divided into three substages, and the nymphs feed in the fi rst substage only (Gill, 1990). These authors affi rmed that since fourth instar nymphs feed only in their fi rst substage, they are more capable of avoiding the effects of neem oil by ingestion. Neem compounds can penetrate into the leaves as observed by Souza & Vendramim (2005), who registered translaminar effect of neem seed extract on B. tabaci nymphs. According to Isman (2006) there are many plant species with antifeedant properties, but among these, only neem seems to cause toxic effects by ingestion, with deleterious physiological consequences to insects, as observed by Bleicher et al. (2007) and Singha et al. (2007).
The results of the present research suggest that neem oil at a 0.5% concentration may be used in preventive applications to reduce initial infestation of B. tabaci on dry bean without possible phytotoxic effects. Moreover, a 0.5% neem oil concentration would be less expensive, considering that one of the major limitations for the use of this product in agriculture is its high refi ning cost (Isman, 2006). Neem oil at a 1% concentration, however, can kill over 80% of the nymphs from fi rst to third instar and 68% of fourth instar nymphs, so it is an alternative for the reduction of late whitefl y infestations. According to Zehnder et al. (2007) arthropods in organic production systems should be managed using multiple strategies, aimed at the reduction of insect pest populations to minimize the need for curative solutions. Based on estimated values of LC 50 and LT 50 for neem oil at 0.5% and 1% concentrations, this botanical extract can be one of those organic production strategies.

Conclusions
1. First, second and third instar nymphs are more susceptible to neem oil than fourth instar nymphs.
2. Neem oil at 1% concentration can reduce the population of fi rst to third nymphal instars by over 80%.
3. Based on estimated values of lethal concentration and lethal time, neem oil at 1% concentration can be effi cient to reduce populations of Bemisia tabaci nymphal stages.